1594
J . Med. Chem. 1989, 32, 1594-1599
Intercalating Agents with Covalent Bond Forming Capability. A Novel Type of Potential Anticancer Agents. 2.' Derivatives of Chrysophanol and Emodin Masao Koyama,' K i y o b u m i Takahashi,? T i n g - C h a o Chou,t Zbigniew Darzynkiewicz,§ Jan Kapuscinski,$ T. Ross Kelly," and Kyoichi A. Watanabe*,t Laboratories of Organic Chemistry, Pharmacology, a n d Experimental Cell Research, Sloan-Kettering Institute for Cancer Research, Memorial Sloan-Kettering Cancer Center, Sloan-Kettering Division of Graduate School of Medical Sciences, Cornell University, New York, New York 10021, and Department of Chemistry, Boston College, Chestnut Hill, Massachusetts 02167. Received July 14, 1988 Fifty-one new C-methyl-modified derivatives of the anthraquinones chrysophanol and emodin or their various methyl ethers were prepared for structure-activity relationship studies of anticancer activity against mouse leukemia L1210 and human leukemia HL-60 cells. Representative compounds were spectrophotometrically studied for their capacity to interact with natural and denatured DNA. In general, those anthraquinones bearing an amino function interact with DNA. l&Dimethoxyanthraquinones are incapable of intercalating into DNA. 1- or 8-Monohydroxymonomethoxyanthraquinones, however, interact with DNA to some extent. No straightforward correlation is apparent between the DNA-affinity data of the compounds studied spectrophotometrically and their cytotoxic effects. Cytotoxic potencies of these compounds on cell growth inhibition during a 72-h period are inversely correlated to their potencies when inhibiting [3H]TdR incorporation into DNA during the initial 30 min of exposure. Surprisingly, some compounds that showed more cytotoxicity did not inhibit initial TdR incorporation (0-30 min), while some others that strongly inhibited T d R incorporation initially did not exhibit cytotoxicity in 72 h. The results suggest that the cytotoxicity produced by these compounds is time dependent and is not a direct result of initial inhibition of DNA replication.
A number of analogues of certain antitumor intercalating agents, s u c h as ellipticine (Figure 1) ,2,3 4'-(9-acridinylamino)methanesulfon-m-aniside(m-AMSA, amsacrine)? and anthracycline antibiotics (e.g., d o x ~ r u b i c i n ) ~h -a ~v e been synthesized i n order to g a i n better t h e r a p e u t i c potential. However, preliminary screening data show that t h e r e is no straightforward structure-activity relationship within each group. These results seem to suggest that although intercalation m a y be a necessary condition, i t may not be sufficient and other factors m a y b e involved that per s e potentiate the a n t i c a n c e r activity. S t u d i e s on the m e c h a n i s m of anticancer action of the indole antibiotic CC10658,9show that it binds to the minor groove of DNA b y nonintercalative means and then slowly alkylates the amino g r o u p of a d e n i n e b y opening the cyclopropane ring i n the antibiotic molecule. W i t h CC1065, covalent binding of the drug w i t h DNA, therefore, seems to be i m p o r t a n t for i t s potent cytotoxic activity. M e r e physical interaction between the d r u g and DNA m a y not be sufficient. Recent s t u d i e s indicate that m-AMSA i n h i b i t s the topoisomerization and c a t e n a t i o n reactions of DNA topoisomerase I1,lo p r o b a b l y by t r a p p i n g the enzyme-DNA
DNA25 are easily a l k y l a t e d b y 9-oxo-NME. T h i s "biooxidative alkylation" has been proposed as a possible (1) Koyama, M.; Kelly, T. R.; Watanabe, K. A. J. Med. Chem. 1988, 31, 283, is considered Part 1 of this series.
(2) Le Pecq, J.-B.; Xuong, N.-D.; Gosse, C.; Paoletti, C. Proc. Natl. Acad. Sci. U.S.A. 1974, 71, 5078. (3) Guthrie, R. W.; Brossi, A.; Mennona, F. A.; Mullin, J. G.; Kierstead, R. W.; Grunberg, E. J . Med. Chem. 1975, 18, 755. (4) Denny, W. A.; Cain, B. F.; Atwell, G. J.; Hansch, C.; Panthananickal, A.; Leo, A. J. Med. Chem. 1982,25, 276. (5) Mosher, C. W.; Wu, H. Y.; Fujiwara, A. N.; Acton, E. M. J . Med. Chem. 1982, 25, 18. (6) Seshadri, R.; Isreal, M.; Pegg, W. J. J. Med. Chem. 1983, 26, 11.
(7) Myers, C. Cancer Chemother. 1986, 8, 52. (8) Chidester, C. G.; Krueger, W. C.; Mizsak, S. A.; Duchamp, D. J.; Martin, D. G. J. Am. Chem. SOC.1981, 103, 7629. (9) Li, L. H.; Swenson, D. H.; Schpok, S. L. F.; Kuentzel, S. L.; Dayton, B. D.; Krueger, W. C. Cancer Res. 1982, 42, 999. (10) Wang, J. C. Annu. Reu. Biochem. 1985,54, 665. (11) Nelson, E. M.; Tewey, K. M.; Liu, L. F. Proc. Natl. Acad. Sci. U.S.A. 1984, 81, 1361. (12) Chen, G. L.; Yang, L.; Rowe, T. C.; Halligan, B. D.; Tewey, K. M.; Liu, L. F. J . Biol. Chem. 1984,259, 13560. (13) Kuhn, K. W.; Pommier, Y.; Kerrigan, D.; Markowitz, J.; Cuvey, complexes.'lJ2 Other substances, such as etoposide (VPJ. M. Natl. Cancer Inst. Monogr. 1987, 4, 61. 16, F i g u r e l ) , adriamycin, and ellipti~ine'~ also stabilize (14) Lesca, P.; Lecointe, P.; Paoletti, C.; Mansuy, D. Biochem. t h e cleavable complex between DNA topoisomerase I1 and Pharmacol. 1977, 26, 2169. (15) Le Pecq, J.-B.; Gosse, C.; Xuong, N. D.; Cros, S.; Paoletti, C. DNA. In the p r e s e n t s t u d y , we show that t h e incorporaCancer Res. 1976, 36, 3067. tion of an alkylating g r o u p into s o m e DNA intercalating (16) Bernadou, J.; Meunier, B.; Meunier, G.; Auclair, C.; Paoletti, a g e n t s greatly e n h a n c e s their antileukemic properties. C. Proc. Natl. Acad. Sci. U.S.A. 1984, 81, 1297. It is well-known that one of the metabolites of ellipticine, (17) Auclair, C.; Paoletti, C. J . Med. Chem. 1981, 24, 289. 9-hydro~yellipticine'~ (9-OH-E), is also a p o t e n t anticancer (18) Bernadou, J.; Meunier, G.; Paoletti, C.; Meunier, B. J. Med. agent.15 2-N-Methyl-9-hydroxyellipticinium (9-OH-NME) Chem. 1983,26, 574. (19) Meunier, G.; Meunier, B.; Auclair, C.; Bernadou, J.; Paoletti, is o n e of t h e m o s t active d r u g s a m o n g the ellipticine C. Tetrahedron Lett. 1983, 26, 574. analogues.16 T h e latter is easily oxidized b y peroxidases (20) Auclair, C.; Voisin, E.; Banoun, H.; Paoletti, C.; Bernadou, J.; to 9-oxo-2-methylellipticinium17JE(9-oxo-NME), which is Meunier, B. J . Med. Chem. 1984, 27, 1161. highly electrophilic and alkylates various n i t r ~ g e n , ' ~ , ~ ~(21) Monsarrat, B.; Maftouh, M.; Meunier, G.; Dugue, B.; Bernasulfur,21v22and oxygenlEJgnucleophiles. Among biological dou, J.; Armand, J. P.; Picard-Fraire, C.; Meunier, B.; Paoletti, macromolecules, proteins,23p ~ l y a d e n y l a t e RNA,24 ,~~ and C. Biochem. Pharmacol. 1983, 32, 3887. (22) Bernadou, J.; Monsarrat, B.; Roche, H.; Armand, J. P.; Paoletti, C.; Meunier, B. Cancer Chemother. Pharmacol. 1985,15, 'Laboratory of Organic Chemistry, Memorial Sloan-Kettering 63. Cancer Center. (23) Auclair, C.; Meunier, B.; Paoletti, C. Biochem. Pharmacol. Laboratory of Pharmacology, Memorial Sloan-Kettering 1984, 32, 3883. Cancer Center. (24) Dugue, B.; Paoletti, C.; Meunier, B. Biochem. Biophys. Res. 8 Laboratory of Experimental Cell Research, Memorial SloanCommun. 1984, 124, 416. Kettering Cancer Center. (25) Auclair, C.; Dugue, B.; Meunier, B.; Paoletti, C. Biochemistry Department of Chemistry, Boston College. 1986, 25, 1240.
*
'
0022-2623/89/1832-1594$01.50/0
0 1989 American Chemical Society
Journal of Medicinal Chemistry, 1989, Vol. 32, No. 7 1595
Potential Anticancer Agents
q3$
U
&
el I i p t it ire
m o c r i n e (n-#w)
0
I a R‘, R2, R 3 = H(chrysophano1) I b R 1 = O H , R 2 . R 3 = H(emodin1 I I a R1= H I R 2 , R3= Me I I b R1=OMe,R 2 , R3=Me
OH
0
p
dH etoposide (VP-16)
I I I a R ‘ = H , R2,R3=Me I I I b R1 = OMe, R 2 , R 3 = Me Va R’= H, R2= H(Me),R3= Me(H) Vb R’=OMe.R2=H(Me).R3=Me(H) VIa R’, R 2 , R 3 = H V I b R1=OMe, R 2 , R 3 = H
doxorlbicin (drionycin)
/I
Ok
n
I
rahelmycin (GlK5)
R ~ O
Figure 1.
mode of anticancer action.26 For other intercalating anticancer agents, such as amsacrine (m-AMSA), and the anthracycline antibiotics, extensive studies on their mechanism of anticancer action11J2.27-31 and QSAR studies directed toward the development of more selective drugs have been conducted.32-34 Whether these intercalators bind covalently to biomolecules has not been established. In order to test our hypothesis that intercalating agents with covalent bond forming capability may exert potent cytocidal activity, we chose chrysophanol Ia and emodin Ib (Figure 2) as the starting materials: Ia and Ib were isolated from crude rhubarb extract.3b Both compounds exhibit little anticancer or cytocidal activity. Their structural aromatic features indicate that these compounds and their derivatives (particularly positively charged ones) may intercalate into double helix of nucleic acids. Chemistry The hydroxy groups at the 1-and 8-positions of I were methylated with methyl sulfate and K2C03 in acetone36 Dugue, B.; Auclair, C.; Meunier, B. Cancer Res. 1986,46,3828. Riou, J. F.; Vilarem, M. J.; Larsen, C. J.; Riou, G. Biochem. Pharmacol. 1986, 35,4409. Crooke, S. T., Reich, S. D., Eds. Anthracyclines: Current Status and New Deuelopments; Academic Press: New York, 1980.
Myers, C. E. In Pharmacologic Principles of Cancer Treatment; Chabner, B., Ed.; Saunders: Philadelphia, 1982; p 416. DiMarco, A. In Cancer Medicine; Holland, J. F., Frei, E., Eds.; Lea and Febiger: Philadelphia, 1982; 2nd ed., p 872. Muggia, F. M., Young, C. W., Carter, S. K. Eds.; Anthracycline Antibiotics in Cancer Therapy; Martinus Nijhoff: Hague/ Boston/New York, 1982; pp 71-174. Ferguson, L. R.; Denny, W. A. J . Med. Chem. 1980,23, 269. Denny, W. A.; Cain, B. F.; Atwell, G. J.; Hansch, C.; Panthananickal, A,; Leo, A. J . Med. Chem. 1982, 25, 276. Alexander, J.; Khanna, I.; Lednicer, D.; Mitscher, L. A.; Veysoelu. T.: Wieloeorski. Z.: Woleemuth. R. L. J . Med. Chem. I
I
.
I
1984,.27,’1343.
Kellv. T. R.: Chandrakumar. N. S.:’ Walters.. N.:. Blancaflor. J. J. 0;g. Chem. 1983, 48, 3573. Alexander, J.; Bhatia, A. V.; Mitscher, L. A.; Omoto, S.; Suzuki, T. J . Org. Chem. 1980, 45, 20.
II
o
I
OR^
IVa R’ = H , R2, R3 = Me I V b R’ = OMe, R 2 , R3 = Me VIIa R’=H, R2=H(Me).R3=Me(H) V I I b R’ =OMe,R2=H(Me),R3=Me(H) VIIIa R 1 , R z , R 3 = H V I I I b R1=OMe, R 2 , R 3 = H 0
IXa R‘ = H , R 2 , R 3 = Me I X b R’ =OMe.R2,R3=Me Xa R ’ = H , R2=H(Me),R3=Me(H) Xb R ’ = O M e , R 2 = H ( M e ) , R 3 = M e ( H ) XIa R’, R 2 , R3 = H X I b R1=OMe. R 2 , R 3 = H
Figure 2. (*) See Table I for NR””.
to the known 1,8-dimethoxy-9,10-anthraquinones 11. The C-methyl group of I1 was then brominated with NBS3‘ or 1,3-dibromo-5,5-dimethylhydantoin (BDH)38,39in carbon tetrachloride in the presence of benzoyl peroxide to give monobromide I11 as the major product along with a small amount of dibromide IV. Treatment of I11 with various amines including mono(2-hydroxyethy1)amineand bis(2hydroxyethy1)amine afforded the corresponding alkylamino derivatives IX (Table I). Chlorination of the (2hydroxyethy1)amino derivatives IX-2 gave the corresponding 3- [ [ (2-chloroethyl)amino]methyl]-9,l0-anthraquinones IX-3. On the basis of reports by Anderson et al.4”3 that certain carbamates are susceptible to nucleo~~
~
Banville, J.; Brassard, P. J . Chem. SOC.,Perkin Trans. 1 , 1976, 613.
Sargent, M. V.; Smith, D. 0. N.; Elix, J. A.; Roffey, P. J. Chem. SOC.,C 1969, 2763. Cava, M. P.; Ahmed, Z.; Benfaremo, N.; Murphy, R. A.; 0’Malley, G. J. Tetrahedron 1984, 40, 4767. Anderson, W. K.; Corey, P. F. J . Med. Chem. 1977,20, 812. Anderson, W. K.; Halat, M. J. J . Med. Chem. 1979, 22, 977. Anderson, W. K.; Chang, C.-P.; McPherson, H. W. J. Med. Chem. 1983, 26, 1333. Zwelling, L. A. Cancer Metastasis Reu. 1985, 4 , 263.
1596 Journal of Medicinal Chemistry, 1989, Vol. 32, No. 7
Koyama et al.
Table 1. Synthetic Derivatives of Chrysophanol and Emodin 0
/ R 3 1 *4
compda IIa IIb IIIa IIIb IVa IVb Va Vb VIa VIb VIIa VIIb VIIIa VIIIb IXa-1 IXa-2 IXa-3 IXa-4 IXa-5 IXa-6 IXa-7 IXb-1 IXb-2 IXb-3 IXb-5 IXb-6 IXb-7 Xa-1 Xa-2 Xa-3 Xa-4 Xb-1 Xb-2 Xb-3 XIa-1 XIa-2 XIa-3 XIa-4 XIa-5 XIa-6 XIa-7 XIa-8 XIa-9 XIa-10 XIa- 11 XIa-12 XIa-13 XIa-14 XIa-15 XIb-1 XIb-2
R'O
o
OF?
R1 H OMe H OMe H OMe H OMe H OMe H OMe H OMe H H H H H H H OMe OMe OMe OMe OMe OMe H H H H OMe OMe OMe H H H H H H H H H H H H H H H OMe OMe
R2 R3 R4 mp, OCbtc formula Me Me Me 189-190 Me Me Me 228-229 Me Me 176-178 CHzBr Me Me CHzBr 250-254 Me Me CHBr, 207-210 Me Me 254-257 CHBr2 H or Me Me or H 213-2 15 CHzBr Me or H H or Me CHzBr 199-201 H H CHzBr 220-222 H H 249-250 CHzBr Me or H H or Me 176-178 CHBrz H or Me Me or H 202-203 CHBr, H H 211-213 CHBr2 H H 237-238 CHBr, Me Me 154-158 CHzNEtz Me Me 202-205d CHzN(CHzCHZ0H)z Me Me 205-206d CHzN(CHZCHzC1)2 Me Me 120d CHzN(CHzCH20CONHMe)z Me Me 254-255d CHzNHEt Me Me 251-252d CHZNHCHZCHZOH Me Me 208-209d CH2NHCH2CH2Cld CH2NHCH2CH&ld - Me Me 222-223d CHiNEt, CHzNEtz Me Me 225-227d CHzN(CH2CHZOH)z Me Me 200-201d CHzN(CHzCH2Cl)2 Me Me CHzNHEt 267-269d Me Me CHZNHCHZCHZOH 252-253d CHzNHCHZCHzOH Me Me 204-205d CH2NHCH2CH2Cl-HCl*DMF CH2NHCH2CH2CI.HCl.DMF H or Me Me or H 225-227d H or Me Me or H 209-216d H or Me Me or H 203-205d H or Me Me or H 178-182d H or Me Me or H 110-112 H or Me Me or H 221-223 H or Me Me or H 152-154 H H 235-238d CHzNEtz H H 204-207d CHzN(CH2CHZOH)2 H H 211-214d CHzN(CHzCHZC1)z H H 125-131 CHzN(CHzCHzOCONHMe)2 H CH,NHEt H >275 H H CH~NHCH~CH~OH 255-261d H H 255-261d CHzNHCH2CH2Cld H H CHZNHZ 240-245d H H 282-283d CHzNMez H H >275 CH,NH(CHAMe H H 251-252d H H 159-160 H H 255-257d H H 246-247d H H 270-272d CHz(imidazol-l-yl) H H 240-241d CHzNEtz H H 225-227d CHzN(CHzCH20H)z H XIb-3 OMe H 203-206d CH2N(CHzCHzCl)Z H XIb-5 OMe H >275 CHzNHEt H XIb-6 H CHzNHCHzCHzOH 259-260d OMe H XIb-7 OMe H powder CH,NHCHpCH,Cl "All the compounds were analyzed for C, H, X (Br or Cl), and/or N. Analyses for these elements were within 10.4% of the theoretical values required unless specifies otherwise. *For nitrogen-containing compounds, melting points were of the HX salt. d = decomposition. Unstable, and satisfactory analyses could not be obtained.
philic attack, we synthesized N-methylcarbamate IX-4 by treatment of IX-2 with N-methyl isocyanate. The methyl protecting groups at 1 and 8 could be removed stepwise at various stages (Figure 2). Thus, treatment of IX with HBr in acetic acid at room ternperature afforded a crystalline mixture of 1-0-methyl- and 8-0-methylanthraquinones X, whereas acid hydrolysis at
reflux temperature for a few hours resulted in complete demethylation, giving rise to XI. Later, it was found that partially methylated chrysophanol and emodin could be directly brominated to the corresponding mixtures of the monobromides V (major products) and dibromides VII. The former were treated with amines to give X, which were further converted to the corresponding XI. Alternatively,
Journal of Medicinal Chemistry, 1989, Vol. 32, No. 7 1597
Potential Anticancer Agents
/
.00‘ . 350
.
400 Wavelength,
\
450
500
nm
Figure 3. Visible light absorption spectrum of drivative XIb-6 (11.3 pM in the buffer containing 0.01 M NaCl) alone (solid line) and in the presence of 0.2 WMcalf thymus DNA (Sigma type 1) (broken line).
the 3-[(alky1amino)methyll derivatives XI were prepared by amination of VI. Chlorination of XIa-2 with thionyl chloride afforded XIa-3. N-Methyl- and N-isopropylcarbamates, XIa-4 and XIa-12, respectively, were prepared by treatment of XIa-2 with the corresponding N-alkyl isocyanates. Spectrophotometric Studies for DNA Interactions Some representative compounds were studied for their ability to interact with nucleic acids in solution by comparison of the electronic spectrum of drug alone with that of the drug in the presence of an excess of nucleic acid. All drugs studied have an absorption band in the visible region, separate from the absorption band of the nucleic acids, and therefore any changes in band intensity and position were indicative of drug chromophore-DNA interaction. It was observed that both chrysophanol (la) and emodin (Ib) and their derivatives lacking the basic center do not appear to interact with DNA to any significant extent. Since these compounds have low solubility in aqueous solutions, the spectral measurement alone may not be sufficient however to allow one to draw a definite conclusion. Generally, those anthraquinones bearing an amino function (e.g., Xa-1, Xb-1, XIa-1, XIa-2, XIb-2, and XIb-6) interact with both native and thermally denatured DNA but more strongly with native DNA (Figure 3). As expected, 1,8-di-O-methylchrysophanoland 1,6,8-tri-Omethylemodin analogues IXa and IXb do not interact with DNA. Anthraquinones wherein one peri-hydroxyl group is methylated (e.g., Xa-2 and Xb-2) interact with DNA to lesser extent than the corresponding unmethylated (XIa-2 and XIb-2). It is interesting to note that while DNA induced changes in absorption spectra of some derivatives (e.g., XIa-7 and XIb-7)) these do not appear to be connected with their alkylating capabilities, in view of the fact that they could be reversed by addition of MezSO (to 1:l v/v). The dissociation of nonbonded ligand-DNA complexes in the presence of organic solvents is a phenomenon well d ~ c u m e n t e d . ~ ~ Changes in the absorption spectra of the drugs (Figure 3) are not inconsistent with the possibility of the intercalative mode of binding. Other types of nonbonding interactions, however, (e.g., binding to the minor groove of the double h e l i ~ ~cannot ~ , ~ ~be) excluded. It is well(44) Wakelin, L. P. G.; Adams, A.; Hunter, C.; Waring, M. J. Biochemistry 1981, 20, 5779. (45) Bloomfield, V. A.; Crothers, D. M.; Tinoco, I. In Physical Chemistry of Nucleic Acids; Harper and Row: New York, 1974; pp 432-434.
Table 11. Biological Activities of Derivatives of Chrysophanol and Emodin IDEQ, IDm ID,, M (L1210 M [HL-60 cell M (HL-60 TdR compd cell growth) growth (72 h)] ( A ) into DNA) ( B ) B / A IIa 2.8 X lo-‘ 1.0 x 10-5 1.9 x 10-5 1.9 IIb 1.0 x 104 4.9 x 10-5 2.1 x 10“ 0.43 2.5 4.4 x 10” 1.1 x 10-5 IIIa 9.2 X 10” 0.14 1.2 x 10-5 IIIb 6.8 X 8.4 x 10-5 9.0 x 10” 2.1 x 10” 4.3 IVa 8.9 X 9.7 IVb 4.1 X 6.0 X 10” 5.8 X 1.7 7.1 X 10“ 1.2 x 10-5 Va 8.8 X 10” Vb 6.8 X 41 1.7 X 10” 7.0 X 7.1 X 10” 2.3 1.6 x 10-5 VIa 8.6 X 10” VIb 5.9 x lo+ 13.6 2.5 X 3.4 x 10-4 1.3 X lo-‘ VIIa 4.2 X lo-” 6.0 X lo-* 217 2.3 x 10-5 VIIb 1.0 X lo* 9.7 x 10-8 237 2.5 X 10” VIIIa 4.4 x lo-’ 3.9 x 10-5 15.6 1.9 x 10-7 1.1 x 10-4 VIIIb 1.0 X 579 IXa-1 1.3 X lo4 3.0 X 1.9 x 10-5 6.3 1.4 X loF6 0.03 4.9 x 104 IXa-2 >5.0 X 1.4 x 10-5 IXa-3 1.3 X 7.8 1.8 X lo4 0.99 IXa-4 7.2 X 9.5 x 10” 9.4 x 10” 2.9 x 10-5 4.8 x 10-5 IXa-5 1.2 X 10” 0.60 4.6 x 10-5 0.84 5.5 x 10-5 IXa-6 1.4 X lo4 8.6 X 0.22 1.9 x 10-5 IXa-7 2.6 X 10“ 7.9 x 10“ 4.2 X 10“ 0.53 IXb-1 6.9 X 10” IXb-2 >2.7 X 1.0 x 10-4 1.7 x 10-5 0.17 2.0 x 10” 4.5 IXb-3 2.7 X 10” 8.9 X 10“ 1.3 X 1.7 X 0.76 IXb-5 9.8 X 10” 1.1 x 104 2.6 x 10-5 0.24 IXb-6 1.7 X 1.8 x 10-5 IXb-7 3.2 X 5.8 X lo4 3.1 1.4 x 10-5 Xa-1 1.2 x io+ 5.2 X 10” 2.7 4.2 Xa-2 >2.4 X 10” 2.1 x 10-5 8.9 X 1.5 x 10-5 3.9 x 10-7 38.5 Xa-3 1.4 X 10” Xa-4 8.9 X loF5 1.2 x 10-5 1.2 x 10-5 1 1.8 X lo4 26.9 Xb-1 6.9 X 10” 6.7 X 10” 9.6 X 10” >5.0 X lo4 0.02 Xb-2 8.8 X 10“ 5.2 x 10-7 10.4 5.4 x 10“ Xb-3 3.3 X 10” 1.8 X 10“ 1.4 x 10-5 XIa-1 2.8 X 10“ 7.8 4.2 XIa-2 5.9 x 10” 3.3 x 10” 1.4 x 10-5 6.9 x 10-5 1.8 x 10-7 XIa-3 1.3 X 383 2.1 x 10-4 2.8 x 10-5 XIa-4 >1.8 X 0.13 1.8 X 2.1 8.7 X 10“ XIa-5 7.7 X 1.5 x 10-5 17.4 XIa-6 1.6 X 8.6 X XIa-7 7.1 X 10” 7.5 x 10” 3.9 x 10-5 5.2 2.5 x 10-5 1.1 x 10-5 2.3 XIa-8 8.5 X 10“ 2.7 X lo4 XIa-9 2.2 x lo6 3.3 9.3 x 10“ 6.1 2.5 X XIa-10 4.6 X 10” 4.1 X 10” 3.2 X lo4 1.6 x 10-5 5.0 XIa-11 6.7 X 1.7 X 2.7 X 0.63 XIa-12 2.7 X 4.8 XIa-13 1.6 X 10” 2.1 x 10” 1.0 x 10-5 6.3 x 10-5 2.8 X 10“ 22.5 XIa-14 4.0 X 10” 4.8 X 10” XIa-15 7.4 X 10” 4.3 x 10“ 9.0 3.4 x 10“ XIb-1 1.2 X 10” 10.3 3.5 x 10-5 4.5 x 10-5 XIb-2 1.4 X 10“ 6.6 X 10“ 5.2 6.1 x 10-7 1.2 x 10-5 19.7 XIb-3 2.3 X 10” 6.2 2.1 x 10“ 1.3 x 10-5 XIb-5 7.2 X 1.6 X 10“ 1.2 x 10-5 XIb-6 5.0 X 7.5 9.2 X 10” 2.2 x 10-4 23.9 XIb-7 1.8 X
known that intercalative binding, which most often has an ionic component, is affected by a rise in concentration of and compounds such as Xa-2, Xb-2, XIb-3, and XIb-7 lost the ability to interact with DNA when Na+ concentration was increased from 0.01 to 0.1 M. On the basis of this fact one can conclude that the affinity for DNA of these compounds is lower than those that interact with DNA at both ionic strengths. No straightforward correlation is apparent between the DNA-affinity data of the drugs studied and their biological activity. (46) Jorgenson, K. F.; Varsheny, U.; van de Sande, J. H. J. Biomol. Struct. Dyn. 1988, 6, 1005. (47) Kapuscinski, J.; Darzynkiewicz, Z. J . Biomol. Struct. Dyn. 1987, 5 , 127.
1598 Journal of Medicinal Chemistry, 1989, VoE. 32, No. 7 Table 111. Inverse Relationship between Cell Growth Inhibition and Inhibition of Initial Thymidine Incorporation into DNA in HL-60 Cells by Chrysophanol Derivatives4 no. Of "50 ratio of IC5o(dThd incorpn)/ of compds growth)$rM ICso (cell growth), mean f SE mean f SE examined ranee 0.36 f 0.10 167.3 f 72.4 8 50 204.4 f 65.4 0.22 f 0.09 "Cell growth inhibition was measured at the end of 72-h exposure to each compound as described under Experimental Secton. Inhibition of [3H]dThd incorporation into DNA was measured during the first 30 min of exposure to each corresponding compound as described under Experimental Section.
*
Biological Activities Preliminary biological data for inhibiting cell growth of murine L1210 leukemic cells and human acute promyelocytic leukemia cells (HL-60) during 72 h of exposure to the compounds are given in Table 11. The potencies for inhibiting [3H]TdR incorporation into DNA in HL-60 cells during the initial 30-min period are also given in Table 11. It is interesting to note that 1,8-di-O-methyl derivatives are uniformly devoid of activity against L1210 leukemic cells. These results are consistent with published data that compare the ratios of the potencies (IDm's) for cell growth inhibition ( A )and inhibition of [3H]TdR incorporation into DNA in HL-60 cells (B). The B / A ratios allow an indirect estimation of whether or not cytotoxicities exerted by these analogues are primarily due to initial inhibition of DNA synthesis. The B / A ratios for Vb, VIIa, VIIb, VIIIb, Xa-3, and XIa-3 are 41, 217, 237,579, 39, and 383, respectively, suggesting that these compounds exert their initial effects mainly on processes other than DNA synthesis per se (Table 111). These results suggest that these compounds exert their cytotoxic effects in a time-dependent manner and their initial action is targeted at the sites other than DNA elongation. Whether these analogues, like m-AMSA, ellipticine, or anthracyclines, act by inhibiting DNA topoisomerase I1 remains to be explored. It is of interest to note that the above-mentioned anthraquinones are among the most potent antileukemic analogues listed in Table 11, with ID,, values ranging from 1.4 X lo4 to 4.2 X 10-l' for L121OMcells and 1.7 X lo4 to 6.0 X lo4 M for HL-60 cells. Our preliminary experiments indicate that the compounds arrest cells in the S and/or G2 phases of the cell cycle (unpublished results). Experimental Section Melting points were determined on a Thomas-Hoover capillary apparatus and me uncorrected. Elemental analyses were performed by M-H-W Laboratories, Phoenix, AZ, and all new compounds with ,the exception of IXa-7 and XIa-7, which were unstable, ekalyzed correctly. 'H NMR spectra were recorded on a JEOL FY9OQ spectrometer with Me4Sias the internal standard. Chrysophanol (Ia) and emodin (Ib) were isolated from rhubarb extract by the procedure of Kelly et al.= except CH2C12was used instead of Et,O throughout the isolation process. 1,8-Dimethoxy-3-methyl-9,lO-anthraquinone (1,8-Di-Omethylchrysophanol, IIa). A mixture of chrysophanol (Ia, 7.0 g, 0.029 mol), KZCO3 (10 g, 0.071 mol), and Me2S04(10 mL, 0.1 mol) in Me2C0 (300 mL) was stirred under reflux for 16 h and then concentrated in vacuo. The residue was triturated well with water (300 mL), and the crystalline IIa (7.5 g, 96%) was collected by filtration and air-dried: mp 191-193 "C (lit.49mp 195 "C); 'H (48) Burchenal, J. H.; Chou, T.-C.; Lokys, L.; Smith, R. S.; Watanabe, K. A.; Su, T.-L.; Fox, J. J. Cancer Res. 1982,42, 2598.
Koyama et al. NMR (CDC13) 6 2.46 (3 H, s, 3-Me), 3.98 (3 H, s, OMe), 3.99 (3 H, s, OMe), 7.08-7.86 ( 5 H, m, H-2,4,5,6,7). Anal. (C17H1404) C, H. 1,3,8-Trimethoxy-6-methyl-9,10-ant hraquinone (1,3,8Tri-0-methylemodin,IIb). In a similar manner, emodin (Ib, 1.0 g, 3.9 mmol) was converted into IIb (1.04 g, 90%): mp 225 "C (lit." mp 225 "C); 'H NMR (MezSO-d6)6 2.43 (3 H, s, 6-Me), 3.88 (3 H, s, OMe), 3.92 (6 H, s, 2 X OMe), 6.94 (1H, d, H-7, J5,7 = 2.2 Hz), 7.13 (1 H, d, H 4 J 5 . 7 = 2.2 Hz), 7.33 (1H, S, H-2), 7.46 (1 H , s, H-4). Anal. (Cl8Hl6O5)C, H.
3-(Bromomethy1)-1,8-dimethoxy-9,10-anthraquinone (IIIa). To a hot solution of IIa (4.5 g, 0.016 mol) and BDH (2.75 g, 0.019 mol) in C C 4 (500 mL) was added Bz202(0.7 g), and the mixture was heated under reflux for 5 h. The mixture was allowed to cool to room temperature. Insoluble hydantoin was removed by filtration, the filtrate was concentrated in vacuo, and the residue was crystallized twice from EtOAc to give IIIa (3.3 g, 57%): mp 176-178 "C (lit.51mp 174-175 "C); 'H NMR (CDClJ 6 4.01 (3 H, s, OMe), 4.03 (3 H, s, OMe), 4.52 (2 H, s, CH,Br), 7.26-7.84 ( 5 H , m, H-2,4,5,6,7). Anal. (C17H13Br04)C, H , Br.
3-(Dibromomethy1)-1,8-dimethoxy-9,10-anthraquinone (IVa). The mother liquors of recrystallization were concentrated, and the residue was chromatographed on a silica gel column using a mixture of CsH6 and EtOAc (3:l) to give IVa (0.48 g): mp 207-210 "C; 'H NMR (CDCl,) 6 4.01 (3 H, s, OMe), 4.07 (3 H, s, OMe), 6.67 (1H, s, CHBr,), 7.26 (5 H, m, H-2,4,5,6,7). Anal. (C17H12Br204) C, H, Br. A further amount of IIIa (0.33 g) was eluted from the column, making the total yield of 62.7 %.
6 4Dibromomethy1)- 1,3,8-trimethoxy-9,1O-anthraquinone (IVb). In a similar manner, IIb (3.12 g, 0.01 mol) was brominated to give IIIb (3.06 g, 74.4%) [mp 250-254 "C (lit.% mp 233.5-234 "C); 'H NMR (CDC13) was identical with that reported36] and IVb (297 mg) [mp 254-257 "C; 'H NMR (CDC13)6 3.95 (6 H, s, 2 X OMe), 3.97 (3 H, s, OMe), 6.67 (1 H, s, CHBr,), 6.78 (1 H, d, H-7,55,7 = 2.47 Hz), 7.32 (1 H, d, H-5), 7.54 (1 H, d, H-2, J 2 , 4 = 1.92 Hz), 7.90 (1 H , d, H-4)]. Anal. (Cl7HI4Br2O5)C, H, Br.
3-(Bromomethy1)-l(and 8)-hydroxy-d(and 1)-methoxy9,lO-anthraquinone(Va). A mixture of IIIa (70 mg, 0.19 mmol) in HOAc (10 mL) and 30% HBr/HOAc (1 mL) was stirred overnight at room temperature and then concentrated in vacuo. The residue was chromatographed on a silica gel column using CHC1, as the eluent to give 51 mg (76%) of Va as yellow crystals: mp 213-215 "C; 'H NMR (CDC1,) 6 4.08, 4.10 (2 X 3 H, 2 s, 1- and 8-OMe), 4.47, 4.54 (2 X 2 H, 2 s, CH,Br), 7.24-8.04 (10 H, m, H-2,4,5,6,7). Anal. (Cl6Hl1BrO,) C, H , Br. 3 4Dibromomet hy1)-1 (and 8)-hydroxy-8(and 1)-methoxy9,lO-anthraquinone (VIIa) and 6 4Dibromomethy1)- 1 (and
8)-hydroxy-3,8(and 1,3)-dimethoxy-9,10-anthraquinone (VIIb). In a similar manner, from IVa (200 mg, 0.453 mmol) and IVb (100 mg, 0.213 mmol), VIIa (82 mg, 42.5%) and VIIb (82 mg, 84.4%), respectively, were prepared (see Table I).
3-(Bromomethyl)-1,8-dihydroxy-9,lO-anthraquinone (VIa). A mixture of IIIa (1.05 g, 2.9 mmol), HOAc (50 mL), and 30% HBr/HOAc (5 mL) was heated at 100 "C for 5 h. After the mixture was cooled, VIa was collected by filtration, washed with HOAc and H 2 0 , and then air-dried to give 879 mg (91%) of the product: mp 220-222 "C; 'H NMR (CDCI,) 6 4.47 (2 H, s, CH2Br), 7.22-7.90 ( 5 H, m, H-2,4,5,6,7), 12.01, 12.04 (2 X H, 2 s, 1-OH, 8-OH). Anal. (C15H9Br04)C, H, Br. 3-(Dibromomethyl)-1,8-dihydroxy-9,10-anthraquinone (VIIIa) and 6-(Dibromomethyl)-3-methoxy-1,8-dihydroxy9,lO-anthraquinone(VIIIb). In a similar manner, IVa (237 mg, 0.54 mmol) and IVb (472 mg, 1 mmol) were converted into VIIIa (166 mg, 75%) and VIIIb (408 mg, 95%), respectively (Table I). 3-[ (Diethylamino)methyl]-1,8-dimethoxy-9,lO-anthraquinone (IXa-1). To a solution of IIIa (200 mg, 0.55 mmol) in DMF (10.0 mL) was added E h N H (5.0 mL), and the mixture was stirred a t room temperature for 3 days. The mixture was partitioned between EtOAc (20 mL) and H 2 0 (20 mL). The aqueous (49) Bedstein, 8, 473. (50) Bedstein, 8, 523. (51) Blankespoor, R. L.; Schutt, D. L.; Tubergen, M. B.; De Jong, R. L. J . Org. Chem. 1987,52, 2059.
Potential Anticancer Agents layer was washed with EtOAc (20 mL). The combined EtOAc solutions were washed with H 2 0 (2 x 20 mL) and saturated NaCl (2 x 20 mL), dried (Na2S04),and concentrated, and the residue was chromatographed on a silica gel column first with CHC13, which eluted the 3-(hydroxymethyl) derivative (33 mg), followed by CHC13 containing 3% MeOH. The IXa (140 mg) that eluted from the column was converted to the crystalline HCl salt (142 mg, 63%), mp 154-158 "C. In a similar manner but by using the corresponding amines, IXa-2, -5, and -6, were prepared (Table I). Also by use of the same procedure starting from IIIb and the corresponding amines, IXb-1, -2, -5, and -7 were synthesized (Table I). 3 4 (Diethy1amino)methyll-1(or 8)-hydroxy-8(or 1)-methoxy-9,lO-anthraquinone (Xa-1). The HCl salt monohydrate of IXa (100 mg, 0.25 mmol) was dissolved in a mixture of HOAc ( 5 mL) and 30% HBr/HOAc (0.4 mL), and the solution was stirred at room temperature for 24 h. After concentration in vacuo, the residue was partitioned between saturated NaHC03 (10 mL) and CHCl, (10 mL). The CHCl, layer was dried (Na2S04)and concentrated and the residue chromatographed on a silica gel column using CHC1,-MeOH (30:l v/v) to give Xa-1 as a glass, which was dissolved in 1 N HBr (1 mL). Upon dilution of the solution with EtOH ( 5 mL), the monohydrobromide of Xa-1 (82 mg) precipitated as yellow microcrystals, mp 225-227 "C dec. In a similar manner, Xa-2-4 and Xb-1-4 were prepared (Table I). 3-[ [N , N -Bis (2-hydroxyet hyl)amino]met hyll- 1,8-dihydroxy-9,lO-anthraquinone(XIa-2). A mixture of VIa (456 mg, 0.73 "01) and bis(2-hydroxyethy1)amine (600 mg, 5.50 mmol) in DMF (20 mL) was stirred for 2 h and then partitioned between CHCl:, (100 mL) and H 2 0 (100 mL). The organic layer was separated, washed ( H 2 0 ,3 x 50 mL), dried (Na2S04),and concentrated in vacuo. The residue was chromatographed on a silica gel column using CHC1,-MeOH (15:l v/v) as the eluent. The major fraction was concentrated and the residue dissolved in 1 N HC1. After concentration of the solution in vacuo, the residue was triturated well with MeOH ( 5 mL), and the crystalline HC1 salt of XIa-2 (496 mg, 91%) was collected by filtration, mp 204-207 "C. Anal. (ClSHl9NO6.HC1)c , H, N. In a similar manner but by use of the corresponding amines, XIa-1, -2, -5, -6, -8-11, and -13-15 were prepared. Also, from VIb, by the same procedure, XIb-1, -2, -5, and -6 were obtained (Table I). 3-[ [ N ,N - B i s (2-~hloroethyl)amino]methyl]1,8-dihydroxy-9,lO-anthraquinone(XIa-3). T o a solution of XIa2.HC1 (1.05 g, 2.57 mmol) in dry DMF (40 mL) was added SOC12 (1.0 mL), and the solution was stirred at room temperature for 1.5 h. The solution was concentrated in vacuo (bath temperature
